The Effects Of Platelet-activating Factor On Pain Modulation In Spinal Cord

BackgroundTissue damage and inflammation usually induce persistant pain, characterized asspontaneous pain, hyperalgesia, tactile allodynia and so on. Within the past decade, there isgreat development in research on persistent pain.Central sensitization in pain signalconduction pathway is viewed as important mechanism in the creation and maintenance ofpersistent pain. However, so far the mechanism of central sensitization remainsunintelligible.Years of research indicates an amount of endogenous nerve mediators areimplicated in pain signal conduction and modulation in central sensitization.Platelet-activating factor (PAF) is an endogenous phospholipid, described in a varietyof immune and inflammatory cells and a potent mediator that participates in inflammatoryresponses.In further researches, neural cells produce PAF and also are target cells of PAF.PAF is suggested to be implicated in a variety of physiological and pathological states incentral nervous system, and play a significant role in signal conduction in nervous system.Recently, evidence suggests that PAF may play a part in pain signal conduction andmodulation in central nervous system. At present, the role and mechanism of PAF acted inpain signal conduction and modulation remain to be established further. The role of prostaglandins (PGs) as nociceptive mediators in inflammatory pain isaccepted extensively. Along with intensive research, the role of prostaglandins in other painmodels such as nuropathic pain is recognized gradually. Prostaglandins are thought to playan important role in the creation and maintenance of persistent pain both at peripheral sitesand in central nervous system. Spinal cord is the primary center in pain signal transmission,and prostaglandins in spinal cord act as important mediators in the creation andmaintenance of persistent pain. In vitro, PAF induces prostaglandin E₂ (PGE₂) release fromrat primary astrocytes.It suggests that prostaglandins and their rate-limiting enzymes (i.e.cyclooxygenases) may be involved in pain modulation mediated by PAF in central nervoussystem.There are no reports corresponding to this subject.Traditionally, glial cells were simply thought to be housekeepers for neurons andprovide neurochemical precursors and energy sources to neurons, not to function in signaltransmission among neural cells, and it were neurons and their transmitters rather than glialcells that modulated and amplified signals from the periphery in spinal cord and wereresponsible for hyperalgesia. However, in the last several years, we have seen exponentiallyincreasing number of research articles suggesting the role of glial cells in pain controlespecially in conditions involving inflammation and nerve injury. Activation of glial cells(e.g. astrocytes and microglia) and proinflammatory cytokines release from glial cellscontribute importantly to the development of pain hypersensitivity under pathologicalconditions. PAF receptor mRNA and functional expression are abundant in spinal cord andglial cells. In addition, PAF receptors function to stimulate the production of transmittersand proinflammatory cytokines in cultured glial cells in vitro.These researches abovesuggest the activation of glial cells and release of proinflammatory cytokines may beimplicated in pain modulation mediated by PAF in central nervous system. There are noreports corresponding to this subject.Evidence suggests that PAF exerts cellular actions through binding to its receptors.PAF has two kinds of binding sites which are high-affinity intracellular membrane binding site and low-affinity cell surface receptor. Inhibition of platelet-activating factor receptorsattenuates the inflammatory nociceptive response in rats.Whether antagonists for PAFreceptors attenuate nociceptive response in other pain model such as neuropathic pain hasnot been reported.To further explore the role of PAF in nociceptive modulation and its mechanism, thepresent study is to investigate: (1) Effects of platelet-activating factor administeredintrathecally on pain behaviour and cyclooxygenase expression in spinal cord in rats,antinociceptive effects of different inhibitors of cyclooxygenases on pain behaviouralresponse induced by PAF. (2) Effects of platelet-activating factor administeredintrathecally on glial activation and proinflammatory cytokines in spinal cord in rats. (3)Antinociceptive effects of PAF receptor antagonists administered intrathecally in rats withspared nerve injury.Methods and Results1. Effects of platelet-activating factor administered intrathecally on painbehaviour and cyclooxygenase expression in spinal cord in ratsMethods Forty-eight Sprague-Dawley rats were randomly divided into six groups, andintrathecal PE-10 catheters were placed in the spinal subarachnoid space of rats: InACSF(artificial cerebral spinal fluid)group, rats were treated with 10μl artificial cerebralspinal fluid intrathecally. In PAF group, rats were treated with 5μg PAF which wasdissolved in 10μl artificial cerebral spinal fluid. In normal saline, SC-560, NS-398 andindomethacin group, all rats were pretreated with 1.5 ml normal saline, SC-560, NS-398and indomethacin intravenously through tail vein 10 minutes prior to intrathecal PAF(5μg,dissolved in 10μl ACSF) injection respectively. The doses of SC-560, NS-398 andindomethacin dissolved in 1.5 ml normal saline were 150μg/kg, 150μg/kg and 300μg/kg respectively. Baseline paw withdrawal mechanical threshold (PWMT) and paw withdrawalthermal latency (PWTL) were measured before intravenous and intrathecal injection. InACSF and PAF group, behavioural tests were processed at 15 minutes after intrathecalinjection and repeated every 15 minutes in the following 5 hours.In normal saline, SC-560,NS-398 and indomethacin group, behavioural tests were processed at 30 minutes afterintrathecal PAF injection.The rats were euthanized immediately after nociceptivebehavioural tests were all over. RT-PCR analysis (in ACSF and PAF group) andradioimmunoassay (in each group) were used to assess the expressions of COX-1, COX-2and COX-3 mRNA and concentrations of PGE₂ in L_4-6 spinal cord respectively.Results (1) Nociceptive behavioural tests: The mean baseline PWMT of rats in thepresent study was 39.75±1.85 g. Intrathecal injection with 5μg PAF induced developmentof tactile allodynia rapidly in PAF group. PWMT of rats in PAF group in different timepoints reduced significantly compared with that in ACSF control group(P＜0.05).Tactileallodynia in rats existed during the whole observation period of 5 hours, and the peak timespersisted about 60 minutes.Intrathecally administered PAF also induced development ofthermal hyperalgesia rapidly, The mean baseline PWTL of rats in the present study was13.04±0.20 s. PWTL of rats in PAF group in different time points reduced significantlycompared with that in ACSF control group(P＜0.05).The time-course of thermalhyperalgesia was similar to that of tactile allodynia.The cyclooxygenase inhibitors, NS-398and indomethacin attenuated tactile allodynia and thermal hyperalgesia induced by PAF.PWMT and PWTL in NS-398 and indomethacin group increased significantly comparedwith that in normal saline control group(P＜0.05). The COX-1 inhibitor, SC-560 had nosignificant effects on both PWMT and PWTL. (2) RT-PCR analysis of cyclooxygenases inspinal cord: COX-1, COX-2 and COX-3 mRNA all expressed constitutively in spinal cordin ACSF group. Both COX-1 and COX-3 mRNA expressed lower than COX-2 mRNA inspinal cord(P＜0.05). PAF increased the expression of COX-2 mRNA(P＜0.05), and had noeffects on the expression of COX-1 and COX-3 mRNA. (3) Radioimmunoassay of PGE₂ in spinal cord: PGE₂ concentration in PAF group increased significantly at 5 hours afterPAF injection compared with that in ACSF group(P＜0.05). Compared with normal salinegroup, NS-398 and indomethacin inhibited the release of PGE₂ induced by PAF in spinalcord (P＜0.05), and the PGE₂ concentration in NS-398 group is lower than that inindomethacin group(P＜0.05). SC-560 did not affect the release of PGE₂ induced by PAF.2. Effects of platelet-activating factor administered intrathecally on glialactivation and proinflammatory cytokine expression in spinal cord in ratsMethods Sixty-four Sprague-Dawley rats were randomly divided into six groups, andintrathecal PE-10 catheters were placed in the spinal subarachnoid space of rats: In ACSFgroup, rats were treated with 10μl artificial cerebral spinal fluid intrathecally. In PAFgroup, rats were treated with 5μg PAF which was dissolved in 10μl artificial cerebralspinal fluid intrathecally. In DMSO control group, SC-514(10 mg/kg)group, SC-514(50mg/kg)group and SC-514(100 mg/kg)group, DMSO(2 ml, the final concentration ofDMSO in normal saline was 0.1%) and SC-514 which was dissolved in 2 ml 0.1% DMSOwere injected intraperitoneally respectively 2 hours prior to intrathecal injection with 5μgPAF. Baseline PWMT and PWTL were measured before intraperitoneal and intrathecalinjection. PWMT and PWTL were measured at 5, 15, 30, 45, 60, 90 and 120 minutes afterPAF was administered intrathecally. The rats were euthanized immediately afternociceptive behavioural tests were all over. The activation of astrocyte and microglia in L_4-6spinal cord were assessed with immunohistochemical staining of glial fibrillary acid protein(GFAP) and OX-42 respectively in ACSF and PAF group. Proinflammatory cytokinesTNF-a and IL-1βin L_4-6 spinal cord were analyzed with ELISA in each group.Results (1) Nociceptive behavioural tests: Intrathecally administered PAF induceddevelopment of tactile allodynia and thermal hyperalgesia rapidly, decreased PWMT andPWTL in rats in PAF group significantly (P＜0.05). Pretreatment with SC-514dose-dependently attenuated nociceptive behavioural response induced by PAF. Both PWMT and PWTL in different time points in SC-514 groups increased significantlycompared with that in DMSO control group(P＜0.05). (2) Immunohistochemical analysisof glial activation: Intrathecally administered PAF activated astrocyte and microglia labeledrespectively with GFAP and OX-42 staining in dorsal horn of gray matter in PAF group.Activating grade of glia is (++) and (-) in PAF and control group respectively. Opticaldensity of GFAP and OX-42 immunoreactive profiles increased significantly in PAF groupcompared with that in control group (P＜0.05). Activated glia distributed throughout thespinal gray matter, and in dorsal horn, activated glia distributed mostly in laminaⅠ～Ⅱ. (3)ELISA analysis of proinflammatory cytokines: Intrathecally admini- stered PAFsignificantly increased the expressions of TNF-a and IL-1βin lumbar spinal cord at 2 hoursafter intrathecal injection (P＜0.05). Pretreatment with SC-514 dose-dependently inhibitedthe increase of TNF-a and IL-1βexpression induced by PAF in spinal cord (P＜0.05).3. Antinociceptive effects of PAF receptor antagonists administeredintrathecally in rats with spared nerve injuryMethods Sixty-four Sprague-Dawley rats were randomly divided into six groups:sham group, SNI group, DMSO (5μl, 0.1% final concentration in normal saline) controlgroup, BN52021(100μg) group, BN50730(100μg) group and BN52021(100μg)+BN50730(100μg) group. BN52021 and BN50730, the antagonists for PAF receptors weredissolved in 5μl DMSO (the final concentration of DMSO in normal saline was 0.1%)respectively. Spared nerve injury neuropathic pain model was set up in all rats except shamgroup. Intrathecal injections with DMSO or antagonists were administered once a day in 7days in DMSO or antagonists groups. PWMT were measured at 7 days in all rats. The ratswere euthanized immediately after nociceptive behavioural tests were all over.Immunohistochemical staining (in each group),and radioimmunoassay (in ACSF and PAFgroup) were used to assess the expressions of c-fos and PAF levels in L_4-6 spinal cordrespectively. Results (1) Nociceptive behavioural tests: Spared nerve injury induced tactileallodynia in rats. PWMT in SNI group reduced significantly compared with that in shamgroup (P＜0.05). The antagonists BN52021 and BN50730 attenuated tactile allodynia in ratswith spared nerve injury. PWMT in three antagonist groups increased, significantlydifferent from that in SNI and DMSO group (P＜0.05). Combination treatment withBN52021 and BN50730 were more potent in antinociceptive effects than treatment withone alone (P＜0.05). (2) Immunohistochemical analysis: The expression of c-fos in spinalcord distributed extensively in laminaⅠ～Ⅴ,mostly in laminaⅠ～Ⅱ.Fos-like immuno-reactivityneurons were present dyeing deep brown on nucleus, with round or ellipticgranule-like product, and not dyeing in cytoplasm.The number of Fos-like immuno-reactivityneurons in ipsilateral dorsal horn in SNI rats increased, significantly differentfrom that in sham group(P＜0.05). The number of Fos-like immunoreactivity neurons indifferent lamina of dorsal horn in spinal cord in three antagonist groups reduced,significantly different from that in SNI and DMSO group(P＜0.05). (3) Radioimmunoassayof PAF in spinal cord: SNI induced PAF release in spinal cord in SNI group rats at 7days. PAF concentration in spinal cord in SNI group was 1.183±0.098 pg/mg, significantlydifferent from 0.571±0.065 pg/mg in sham group(P＜0.05).Conclusions1. Intrathecally administered PAF may induce development of tactile allodynia andthermal hyperalgesia in rats.The activation and increased expression of COX-2 andrelease of PGE₂ in spinal cord are implicated in its mechanism.2. Intrathecally administered PAF may induce development of tactile allodynia andthermal hyperalgesia in rats.The activation of glia and NF-κB pathway, and theincreased expressions of TNF-αand IL-1βin spinal cord in rats are implicated in its mechanism. IKKβplays the major role in the activation of NF-κB pathway andexpressions of proinflammatory cytokine TNF-αand IL-1βinduced by PAF.3. These findings suggest a role for endogenous PAF in central nociceptive modulationfor spared nerve injury induced neuropathic pain in rats.Both intracellular and cellsurface PAF binding sites are involved in nociceptive transmission and modulation inrats, and that PAF receptor antagonists might be useful for treating some patients withneuropathic pain.